UAV plant image Classification Using Combined Machine Learning And Deep Learning Models

Plant identification using Unmanned aerial vehicles is a powerful tool that can provide farmers and researchers with valuable information about crops and vegetation. In precision agriculture, identifying and classifying plants is a complex process. Due to various factors, such as the quality and resolution of the image, recognizing the exact type can be challenging. For plant identification purposes, more deep learning (DL) and machine learning (ML) methods are used. In this work, an integrated approach combining deep learning (DL) and machine learning (ML) methods is developed to improve the plant-type classification by improving the quality of the image dataset using image preprocessed techniques Enhanced super-resolution generative adversarial network (ESRGAN) and Contrast Limited Adaptive Histogram Equalization (CLAHE). A convolutional neural network 16 layers deep (VGG-16 mode) is used for feature extraction and classification using Machine Learning models including Random Forest classification and Support Vector Machines. To identify the best predictive model, a comparative study was carried and the hybrid method VGG-16 with Support Vector Machine using image preprocessed ESRGAN and CLAHE gives the optimum results of 98.06% accuracy.